Chapter CIV: Part VII: The Dialects of the English Language (4)

[65] It is almost unnecessary to state that the sentence quoted in the text is really a beautiful couplet of Withers's poetry _transposed_. It was advisable to do this, for the sake of guarding against the effect of the rhyme. To have written,

What care I how fair she _is_
If she be not fair to me?

would have made the grammar seem worse than it really was, by disappointing the reader of a rhyme. On the other hand, to have written,

What care I how fair she _were_,
If she were not kind as _fair_?

would have made the grammar seem better than it really was, by supplying one.

[66] In the first edition of the present work I inaccurately stated that _neither_ should take a plural and _either_ a singular verb; adding that "in predicating something concerning _neither you nor I_, a negative assertion is made concerning _both_. In predicating something concerning _either you or I_, a positive assertion is made concerning _one of two_." This Mr. Connon (p. 129) has truly stated to be at variance with the principles laid down by me elsewhere.

[67] Latin Prose Composition, p. 123.

[68] Quoted from Guest's English Rhythms.

[69] To the definition in the text, words like _old_ and _bold_ form no exception. At the first view it may be objected that in words like _old_ there is no part preceding the vowel. Compared, however, with _bold_, the negation of that part constitutes a difference. The same applies to words like _go_ and _lo_, where the negation of a part following the vowel is a point of identity. Furthermore, I may observe, that the word _part_ is used in the singular number. The assertion is not that every individual sound preceding the vowel must be different, but that the aggregate of them must be so. Hence, _pray_ and _bray_ (where the _r_ is common to both forms) form as true a rhyme as _bray_ and _play_, where all the sounds preceding _a_, differ.

[70] For _pros['o]pa_. The Greek has been transliterated into English for the sake of showing the effect of the accents more conveniently.

[71] For the sake of showing the extent to which the _accentual element_ must be recognised in the classical metres, I reprint the following paper On the Doctrine of the Caesura in the Greek senarius, from the Transactions of the Philological Society, June 23, 1843:--

"In respect to the caesura of the Greek tragic senarius, the rules, as laid down by Porson in the Supplement to his Preface to the Hecuba, and as recognized, more or less, by the English school of critics, seem capable of a more general expression, and, at the same time, liable to certain limitations in regard to fact. This becomes apparent when we investigate the principle that serves as the foundation to these rules; in other words, when we exhibit the _rationale_, or doctrine, of the caesura in question. At this we can arrive by taking cognizance of a second element of metre beyond that of quantity.

"It is assumed that the element in metre which goes, in works of different writers, under the name of ictus metricus, or of arsis, is the same as accent, _in the sense of that word in English_. It is this that constitutes the difference between words like _t['y]rant_ and _res['u]me_, or _s['u]rvey_ and _surv['e]y_; or (to take more convenient examples) between the word _A['u]gust_, used as the name of a month, and _aug['u]st_, used as an adjective. Without inquiring how far this coincides with the accent and accentuation of the classical grammarians, it may be stated that, in the forthcoming pages, arsis, ictus metricus, and accent (_in the English sense of the word_), mean one and the same thing. With this view of the arsis, or ictus, we may ask how far, in each particular foot of the senarius, it coincides with the quantity.

_First Foot._--In the first place of a tragic senarius it is a matter of indifference whether the arsis fall on the first or second syllable; that is, it is a matter of indifference whether the foot be sounded as _t['y]rant_ or as _res['u]me_, as _A['u]gust_ or as _aug['u]st_. In the following lines the words [Greek: heko], [Greek: palai], [Greek: eiper], [Greek: tinas], may be pronounced either as [Greek: he'ko], [Greek: pa'lai], [Greek: ei'per], [Greek: ti'nas], or as [Greek: heko'], [Greek: palai'], [Greek: eiper'], [Greek: tina's], without any detriment to the character of the line wherein they occur.

[Greek: He'ko nekron keuthmona kai skotou pulas.]
[Greek: Pa'lai kunegetounta kai metroumenon.]
[Greek: Ei'per dikaios esth' emos ta patrothen.]
[Greek: Ti'nas poth' hedras tasde moi thoazete.]

or,

[Greek: Heko' nekron keuthmona kai skotou pulas.]
[Greek: Palai' kunegetounta kai metroumenon.]
[Greek: Eiper' dikaios esth' emos ta patrothen.]
[Greek: Tina's poth' hedras tasde moi thoazete.]

_Second Foot._--In the second place, it is also a matter of indifference whether the foot be sounded as _A['u]gust_ or as _aug['u]st_. In the first of the four lines quoted above we may say either [Greek: ne'kron] or [Greek: nekro'n], without violating the rhythm of the verse.

_Third Foot._--In this part of the senarius it is no longer a matter of indifference whether the foot be sounded as _A['u]gust_ or as _aug['u]st_; that is, it is no longer a matter of indifference whether the arsis and the quantity coincide. In the circumstance that the last syllable of the third foot _must_ be accented (in the English sense of the word), taken along with a second fact, soon about to be exhibited, lies the doctrine of the penthimimer and hepthimimer caesuras.

The proof of the coincidence between the arsis and the quantity in the third foot is derived partly from _a posteriori_, partly from _a priori_ evidence.

1. In the Supplices of Aeschylus, the Persae, and the Bacchae, three dramas where licences in regard to metre are pre-eminently common, the number of lines wherein the sixth syllable (_i. e._, the last half of the third foot) is without an arsis, is at the highest sixteen, at the lowest five; whilst in the remainder of the extant dramas the proportion is undoubtedly smaller.

2. In all lines where the sixth syllable is destitute of ictus, the iambic character is violated: as

[Greek: Threken perasa'ntes mogis polloi ponoi.]
[Greek: Duoin gerontoi'n de strategeitai phuge.]

These are facts which may be verified either by referring to the tragedians, or by constructing senarii like the lines last quoted. The only difficulty that occurs arises in determining, in a dead language like the Greek, the absence or presence of the arsis. In this matter the writer had satisfied himself of the truth of the two following propositions:--1. That the accentuation of the grammarians denotes some modification of pronunciation other than that which constitutes the difference between _A['u]gust_ and _aug['u]st_; since, if it were not so, the word [Greek: angelon] would be sounded like _m['e]rrily_, and the word [Greek: angelon] like _dis['a]ble_; which is improbable, 2. That the arsis lies upon radical rather than inflectional syllables, and out of two inflectional syllables upon the first rather than the second; as [Greek: ble'p-o, bleps-a's-a], not [Greek: blep-o', bleps-as-a']. The evidence upon these points is derived from the structure of language in general. The _onus probandi_ lies with the author who presumes an arsis (accent in the English sense) on a _non_-radical syllable. Doubts, however, as to the pronunciation of certain words, leave the precise number of lines violating the rule given above undetermined. It is considered sufficient to show that wherever they occur the iambic character is violated.

The circumstance, however, of the last half of the third foot requiring an arsis, brings us only half way towards the doctrine of the caesura. With this must be combined a second fact, arising out of the constitution of the Greek language in respect to its accent. In accordance with the views just exhibited, the author conceives that no Greek word has an arsis upon the last syllable, except in the three following cases:--

1. Monosyllables, not enclitic; as [Greek: spho'n, pa's, chtho'n, dmo's, no'n, nu'n], &c.

2. Circumflex futures; as [Greek: nemo', temo'], &c.

3. Words abbreviated by apocope; in which case the penultimate is converted into a final syllable; [Greek: do'm', pheides'th', kentei't', ego'g'], &c.

Now the fact of a syllable with an arsis being, in Greek, rarely final, taken along with that of the sixth syllable requiring, in the senarius, an arsis, gives as a matter of necessity, the circumstance that, in the Greek drama, the sixth syllable shall occur anywhere rather than at the end of a word; and this is only another way of saying, that, in a tragic senarius, the syllable in question shall generally be followed by other syllables in the same word. All this the author considers as so truly a matter of necessity, that the objection to his view of the Greek caesura must lie either against his idea of the nature of the accents, or nowhere; since, that being admitted, the rest follows of course.

As the sixth syllable must not be final, it must be followed in the same word by one syllable, or by more than one.

1. _The sixth syllable followed by one syllable in the same word._--This is only another name for the seventh syllable occurring at the end of a word, and it gives at once the hepthimimer caesura: as

[Greek: Heko nekron keuthmo'na kai skotou pulas.]
[Greek: Hikteriois kladoi'sin exestemmenoi.]
[Greek: Homou te paiano'n te kai stenagmaton.]

2. _The sixth syllables followed by two_ (_or more_) _syllables in the same word_. This is only another name for the eighth (or some syllable after the eighth) syllable occurring at the end of a word; as

[Greek: Odme broteion hai'maton me prosgela.]
[Greek: Lamprous dunastas em'prepontas aitheri.]

Now this arrangement of syllables, taken by itself, gives anything rather than a hepthimimer; so that if it was at this point that our investigations terminated, little would be done towards the evolution of the _rationale_ of the caesura. It will appear, however, that in those cases where the circumstance of the sixth syllable being followed by two others in the same words, causes the eighth (or some syllable after the eighth) to be final, either a penthimimer caesura, or an equivalent, will, with but few exceptions, be the result. This we may prove by taking the eighth syllable and counting back from it. What _follows_ this syllable is immaterial: it is the number of syllables in the same word that _precedes_ it that demands attention.

1. _The eighth syllable preceded in the same word by nothing._--This is equivalent to the seventh syllable at the end of the preceding word: a state of things which, as noticed above, gives the hepthimimer caesura.

[Greek: Anerithmon gela'sma pam|metor de ge.]

2. _The eighth syllable preceded in the same word by one syllable._--This is equivalent to the sixth syllable at the end of the word preceding; a state of things which, as noticed above, rarely occurs. When however it does occur, one of the three conditions under which a final syllable can take an arsis must accompany it. Each of these conditions requires notice.

[alpha]). With a non-enclitic _mono_-syllable the result is a penthimimer caesura; since the syllable preceding a monosyllable is necessarily final.

[Greek: Heko sebi'zon so'n Klu'tai|mnestra kratos.]

No remark has been made by critics upon lines constructed in this manner, since the caesura is a penthimimer, and consequently their rules are undisturbed.

[beta]). With _poly_-syllabic circumflex futures constituting the third foot, there would be a violation of the current rules respecting the caesura. Notwithstanding this, if the views of the present paper be true, there would be no violation of the iambic character of the senarius. Against such a line as

[Greek: Kago to son nemo' pothei|non aulion]

there is no argument _a priori_ on the score of the iambic character being violated; whilst in respect to objections derived from evidence _a posteriori_, there is sufficient reason for such lines being rare.

[gamma]). With _poly_-syllables abbreviated by apocope, we have the state of things which the metrists have recognised under the name of quasi-caesura; as

[Greek: Kenteite me pheide'sth' ego | 'tekon Parin.]

3. _The eighth syllable preceded in the same word by two syllables._--This is equivalent to the fifth syllable occurring at the end of the word preceding: a state of things which gives the penthimimer caesura; as

[Greek: Odme broteion hai' maton | me prosgela.]
[Greek: Lamprous dunastas em'prepon tas aitheri.]
[Greek: Apsuchon eiko pro'sgeloisa somatos.]

4. _The eighth syllable preceded in the same word by three or more than three syllables._--This is equivalent to the fourth (or some syllable preceding the fourth) syllable occurring at the end of the word preceding; a state of things which would include the third and fourth feet in one and the same word. This concurrence is denounced in the Supplement to the Preface to the Hecuba; where, however, the rule, as in the case of the quasi-caesura, from being based upon merely empirical evidence, requires limitation. In lines like

[Greek: Kai talla poll' epei'kasai | dikaion en,]

or (an imaginary example),

[Greek: Tois soisin aspide'strophois|in andrasi,]

there is no violation of the iambic character, and consequently no reason against similar lines having been written; although from the average proportion of Greek words like [Greek: epeikasai] and [Greek: aspidestrophoisin], there is every reason for their being rare.

After the details just given, the recapitulation is brief.

1. It was essential to the character of the senarius that the sixth syllable, or latter half of the third foot, should have an arsis, ictus metricus, or accent in the English sense. To this condition of the iambic rhythm the Greek tragedians, either consciously or unconsciously, adhered.

2. It was the character of the Greek language to admit an arsis on the last syllable of a word only under circumstances comparatively rare.

3. These two facts, taken together, caused the sixth syllable of a line to be anywhere rather than at the end of a word.

4. If followed by a single syllable in the same word, the result was a hepthimimer caesura.

5. If followed by more syllables than one, some syllable in an earlier part of the line ended the word preceding, and so caused either a penthimimer, a quasi-caesura, or the occurrence of the third and fourth foot in the same word.

6. As these two last-mentioned circumstances were rare, the general phaenomenon presented in the Greek senarius was the occurrence of either the penthimimer or hepthimimer.

7. Respecting these two sorts of caesura, the rules, instead of being exhibited in detail, may be replaced by the simple assertion that there should be an arsis on the sixth syllable. From this the rest follows.

8. Respecting the non-occurrence of the third and fourth feet in the same word, the assertion may be withdrawn entirely.

9. Respecting the quasi-caesura, the rules, if not altogether withdrawn, may be extended to the admission of the last syllable of circumflex futures (or to any other polysyllables with an equal claim to be considered accented on the last syllable) in the latter half of the third foot.

[72] _Sceolon_, _aron_, and a few similar words, are no real exceptions, being in structure not present tenses but preterites.

[73] Quarterly Review, No. clxiv.

[74] Quarterly Review, No. clxiv.

[75] From the Quarterly Review, No. cx.

[76] From the Quarterly Review, No. cx.

[77] Apparently a _lapsus calami_ for _spede_.

[78] J. M. Kemble, "On Anglo-Saxon Runes," _Archaeologia_, vol. xxviii.

[79] But not of _Great Britain_. The Lowland Scotch is, probably, more Danish than any South-British dialect.

[80] In opposition to the typical Northumbrian.

[81] Quarterly Review--_ut supra_.

[82] The subject is a Lincolnshire tradition; the language, also, is pre-eminently Danish. On the other hand, the modern Lincolnshire dialect is by no means evidently descended from it.

[83] For some few details see Phil. Trans., No. 36.

[84] Transactions of the Philological Society. No. 93.

[85] Philological Transactions. No. 84.

[86] Transactions of the Philological Society, No. 92.

[87] Quarterly Review, vol. xliii.

* * * * *

Changes made against printed original.

Page xxv. "227. The combination _-pth_": 'combinations' in original.

Page xxxiv. "465, 466. The Slavonic praeterite": 'paerterite' in original.

Page xli. "676, 677. Rhyme--its parts": '677, 677' in original.

Page 3, s. 9. "The south-eastern parts of Scotland": 'south-western' in original (compare 'south-eastern', 2 sentences earlier).

Page 6, s. 13(3c). "half a century earlier than the epoch of Hengist": 'earlier that' in original.

Page 50. s. 94. "certain Anglo-Saxon inflections.": 'Anglo-Saxons' in original.

Ibid. "h['e]r, d['e]de, br['e]da, Frisian;": 'Frisian; Fris.' in original.

Ibid. "ju=y or eo": 'eo' omitted in original.

Page 71. s. 127. "a population originating in places": 'orginating' in original.

Page 112. s. 174. "Smidhum however, is a single": 'Smdhium' in original.

Page 143. s. 198. "Concerning the consonants as a class": 'vowels' (for 'consonants') in original.

Page 150. s. 212. Table, first row, Lene Flat: "b": 'v' in original (compare s. 203).

Page 158. s. 227. "the th is a (so-called) aspirate": 'the f' in original.

Ibid. "the second may be accommodated to the first, tupt": 'tuft' in original.

Page 160. s. 229. "dh to d": 'th to d' in original.

Page 161. s. 231. "the v in fever": 'the e' in original.

Page 194. s. 255. "the statement ... that ... the c is mute": 'the k' in original.

Page 202. s. 258. "17. Pe Pi.": '17. Pi Phi.' in original.

Page 265. s. 315. "se scearpeste sweord": 'sword' in original.

Page 286. s. 340. "In Anglo-Saxon the termination -ing": 'terminations' in original.

Page 300. s. 355. "I ate ... we ate": 'ete' for 'ate' (twice) in original.

Page 301. Ibid. "swungon, we swung": 'swangon' in original (does not fit criterion for this table).

Page 323. s. 382. "accounting for the -s in must": 'in most' in original.

Page 324. s. 382. "wit, wot, wiss, wist": 'wit, wot, wiss, wsst' in original.

Page 356. s. 411. "the word rose prefixed": 'the word tree prefixed' in original (the same as the contrary case).

Page 368. s. 426(II). "form another order": 'from another order' in original.

Page 398. s. 479. "the words Roman emperor might be wholly ejected": 'the word' in original.

Page 411. s. 507. "in the indicative and subjunctive moods": 'is the' in original.

Page 434. s. 545. "the analogy between the words there and it": 'these and it' in original.

Page 465. s. 581. "will be taken up in p. 475": 's. 475' in original.

Page 482. s. 606. "a pair of propositions connected by the conjunction": 'prepositions' in original.

Page 490. s. 617. "4. Let tupsaimi be considered an aorist subjunctive": 'on aorist' in original.

Page 562. s. 709. "distinguished by their origin only": 'distinguised' in original.

Footnote 8. "the Anglo-Saxon adjectives in s. 85": 's. 20' in original.

Footnote 63. "deciding its true nature": 'rue nature' in original.

End of Project Gutenberg's The English Language, by Robert Gordon Latham

<pre id="pg-footer">